Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Master #7

Closed
wants to merge 11 commits into from
Closed

Master #7

wants to merge 11 commits into from

Conversation

antimodular
Copy link

This modified version does has a face tracking and a pedestrian tracking example. Both examples only run at very low frame rates at this point.

@genekogan
Copy link
Collaborator

thanks! i just tried these out, look very good. i would maybe just want to clean up the code a bit.

the face-tracking example though does run kind of slow, have you compared it to ofxFaceTracker? wondering if it's more accurate so more suitable for non-real-time detection...

@antimodular
Copy link
Author

i certainly know and have worked with ofxFaceTracker before, which works well.
since ofxccv runs so slow i have not compared it a lot to ofxFaceTracker. my sense is that it can detect smaller faces, but is not a lot better for rotated faces.

let me know what part of the code you think needs cleaning and if you want me to take some of this on or give it your own twist.

@genekogan
Copy link
Collaborator

i think just getting rid of the commented/unused code to keep things clean, also i think we'll want to add the necessary data files (face.sqlite and the pedestrian file) to the download script.

also, about facetracking more generally, there's a thread on ofxFaceTracker about using the latest dlib which seems to have super fast and robust facetracking. ofxLearn is built on top of dlib (older version) so i've been meaning to try, just haven't made the time yet. check out the links cited in the first thread -- really impressive stuff.

@antimodular
Copy link
Author

my new commit to my github has the files included. not sure if i have to post a new pull request?

i can go over the code again and remove all unnecessary commands and old code, in the next couple of days.

also, we have ofxDLib working with face tracking example:
https://github.com/roymacdonald/ofxDLib

@genekogan
Copy link
Collaborator

oh i see yeah. they are small enough to package with the library.

wow, i didn't see ofxDLib. looks great! but i don't think it's using the same implementation as the one referenced above, and to my eyes doesn't track as well as ofxFaceTracker.

let's let kyle comment, and we can get to merging.

@kylemcdonald
Copy link
Owner

this generally looks good, but it will need some cleaning. if either of you want to go for it, @genekogan you should feel free to merge it after:

  • the comments are cleaned up
  • readme is cleaned up
  • commits are squashed to one commit (optional, but helpful)
  • remove #ifdef USE_VIDEO parts, this is usually left as exercise for user
  • screenshots are either converted to jpg (to avoid unnecessary repo bloat) or removed completely (better). adding a screenshot to the PR is better than adding it to the repo.

@antimodular
Copy link
Author

-cleaned the comments out
-cleaned the read me
-don't know how to squash things
-removed #ifdef
-converted screenshots to jpg

@kylemcdonald
Copy link
Owner

thanks! i'll take care of the rest.

@kylemcdonald
Copy link
Owner

i just merged it, but unfortunately the easiest way was with a single commit that does not include you in the history. i added antimodular to the readme and in the commit message, let me know if there's anything else i can do. thanks for the PR!

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

4 participants